Cancers are heterogeneous in biology among patients, tumors in the same patient, and within tumors. As a result, they respond differently to therapy per patient, per tumor and within tumors. Different radiotracers and imaging modalities provide information about different aspects of biology and the physio-metabolic environments of the cancer. As a result, a single modality or radiotracer may not provide sufficient information to predict or assess response to therapy. We hypothesize that improved prediction and assessment of response can thus be obtained by combining quantitative image-derived parameters obtained from multiple imaging modalities or radiotracers. We propose to develop, optimize, and validate approaches for combining multiple image-derived parameters obtained from quantitative imaging procedures in order to optimally predict and assess treatment response. In particular, we propose to combine quantitative metrics from PET/CT, SPECT/CT, and MRI. We will first individually optimize the protocols, acquisition parameters, and imaging methods in order to get the most accurate and reliable parameters to combine. Optimally combining the parameters from different modalities requires knowledge of the reproducibility (precision) of the individual quantitative imaging parameters. We will thus use literature search, phantom studies, realistic simulations, and repeated patient studies to characterize the accuracy and precision of the individual quantitative imaging methods. We will then develop methods to combine the metrics to predict or assess treatment response per patient, per tumor and intra-tumor. We will apply and evaluate these methods in three clinical trials: dynamic and static FDG and FIT PET/CT to assess lung cancer response to cytotoxic chemotherapy; PET/CT and DCE- and DW-MRI in breast cancer response; and SPECT/CT, PET/CT and DCE- and DW-MRI to predict response of brain tumors to anti-angiogenic therapy. In these trials imaging parameters and their signatures will be linked to histology or survival outcomes to provide validation of the combined imaging parameter metrics.